Environ. Sci. Technol. 2007, 41, 1921-1927
Hydrogarnet: A Host Phase for Cr(VI) in Chromite Ore Processing Residue (COPR) and Other High pH Wastes S . H I L L I E R , * ,† D . G . L U M S D O N , † R. BRYDSON,‡ AND E. PATERSON† Macaulay Institute, Craigiebuckler, Aberdeen AB15 8QH, UK, and Leeds Electron Microscopy and Spectroscopy Centre, Institute for Materials Research, SPEME, University of Leeds, LS2 9JT, UK
For understanding both the environmental behavior and developing remediation treatments for chromium ore processing residue (COPR) it is important to identify all the potentially soluble sources of Cr(VI). Hydrogarnet has been identified as a major phase in COPR and it has been previously speculated that it has a capacity to host Cr(VI). Here we provide direct evidence of this capacity by demonstrating the incorporation of Cr(VI) into laboratory synthesized hydrogarnet. Electron microscopy and energy dispersive X-ray microanalysis show incorporation of approximately 17 000-22 000 mg Cr(VI) kg-1 hydrogarnet. X-ray powder diffraction data show that peak intensities are altered by chromium substitution and that chromium substituted hydrogarnets have a smaller unit cell than the pure Ca-Al end member. This is consistent with substitution of hydroxyl tetrahedra by smaller chromate tetrahedra. Electron energy loss spectroscopy confirms the tetrahedral coordination and hexavalent oxidation state of chromium in the hydrogarnets. The maximum amount of hexavalent chromium that can be introduced synthetically corresponds to a replacement of about one out of every eight hydroxyl tetrahedral per unit cell by a CrO42tetrahedra and tallies closely with the amount of chromium measured in hydrogarnets from COPR. Chromium-bearing hydrogarnet is the most abundant crystalline phase in millions of tons of COPR contaminating land around Glasgow, Scotland, and was recently identified in COPR from sites in North America. Calculations based on its abundance and its Cr(VI) content indicate that hydrogarnet can host as much as 50% of the Cr(VI) found in some COPR samples.
Introduction In previous work on the solid-state speciation of chromium in chromite ore processing residue (COPR) we have shown that chromium is associated with several crystalline mineral phases (1) known to have contrasting physicochemical properties. For Cr(VI), one of the minerals implicated was a hydrocalumite type phase, a calcium/aluminum layered double hydroxide, known to have anion exchange properties and subsequently shown to exchange chromate with sulfate * Corresponding author phone: +44 (0)1224 318611; fax: +44 (0)1224 311556; e-mail:
[email protected]. † Macaulay Institute. ‡ University of Leeds. 10.1021/es0621997 CCC: $37.00 Published on Web 02/13/2007
2007 American Chemical Society
in laboratory experiments (2). Calculations based on quantitative mineralogical analysis of COPR and hydrocalumite compositions from sites in Scotland (1) indicate that around 4000 mg kg-1 Cr(VI) is held in an exchangeable form in hydrocalumite. However, observations also suggested (1) that a similar amount of Cr(VI) was associated with hydrogarnet, a calcium aluminate common in many high pH wastes, which has no known anion exchange properties. It was speculated that Cr(VI) was structurally incorporated in hydrogarnet (1). The garnet structure consists of alternating tetrahedra and octahedra which share corners to form a three-dimensional network. In classic rock-forming garnets the tetrahedra are silica SiO44- tetrahedra, whereas in hydrogarnet the tetrahedra are composed of four hydroxyls, H4O44-. Octahedral sites in garnets are occupied by trivalent cations such as aluminum in 6-fold coordination. The cavities formed between the corner-sharing tetrahedra and octahedra have the shape of distorted cubes and contain divalent cations, such as calcium in 8-fold coordination. As an example, the end member hydrogarnet, known by its mineralogical name of katoite, has an ideal formula unit of Ca3Al2(H4O4)3 and there are eight such formula units per unit cell (3). Hydrogarnet is a common phase observed in high-alumina cements (4) and in the cement literature this phase is written in a shorthand notation, C3AH6, denoting its oxide molar ratio, i.e., 3 mol of CaO, 1 mol of Al2O3, and 6 mol of H2O. In addition to its occurrence in cements, hydrogarnet occurs in a wide variety of other high pH environments and industrial waste streams. In COPR in Scotland it typically occurs as the main crystalline phase averaging around 25% by weight (1) and its presence has been recently confirmed (up to around 18 wt %) in COPR from New Jersey (5) in the United States. Hydrogarnet is also found in “muds” produced during the Bayer process (6), blast furnace slags (7), and lime-treated fly ashes (8), as well as in numerous situations where cement is used in the immobilization and containment of toxic wastes, including radioactive waste repositories (9). In all of these systems a variety of toxic metals and metalloids, including chromium, are present as oxyanions. As a result there have been numerous studies devoted to elucidating the mechanisms that affect the mobility of Cr(VI) in cementbased matrices (10-15). In this work we describe the laboratory synthesis of hydrogarnet in the presence of chromate and provide evidence for the incorporation of hexavalent chromium into the synthesized hydrogarnet phase. The results are clearly important because, although hydrogarnet is very stable in high pH environments and thus may immobilize Cr(VI), it dissolves readily as pH levels drop and then becomes a source of any Cr(VI) incorporated in its structure. This is of obvious importance in relation to efforts to model and understand the mobility of Cr(VI) in a wide variety of high pH systems where hydrogarnet is likely to form, as well as informing technologies designed to remediate land contaminated with COPR.
Materials and Methods Synthesis of Hydrogarnet. Calcium aluminum hydrogarnets were synthesized using an appropriate stoichiometric mixture of CaO (Alfa Aesar 010923) and Al2O3 γ-alumina (BDH 100084D). The reagents were carefully weighed, ground, and mixed in an agate mortar and pestle, sealed in Teflon bombs with 20 mL of deionized water, and reacted at 150 °C in an oven for 6 days. Replicate experiments were performed in which chromate was added to the system as CaCrO4 (Alfa Aesar 043333). For experiments made with CaCrO4, the quantity of CaO was adjusted accordingly so as to maintain VOL. 41, NO. 6, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1921
the Ca/Al molar ratio (3/2) appropriate for pure hydrogarnet. Three different chromate concentrations were tested: one corresponding to the stoichiometric proportion of chromate required to replace one in every 24 hydroxyl tetrahedra in the hydrogarnet unit cell, the second corresponding to a substitution of one in 12, and the third corresponding to a one in 3 replacement. The first of these concentrations was chosen because it corresponded approximately with the amount of chromium substitution observed in hydrogarnets found in the COPR studied by Hillier et al., (1), while the second and third concentrations were chosen so as to investigate the upper solubility limit of Cr(VI) in hydrogarnet. For brevity the chromium runs are hereafter referred to as 1/24, 1/12, and 1/3. Altogether, twelve runs were made, three without chromium, four each at the 1/24 and 1/12 concentrations, and one at the 1/3 concentration. Scanning Electron Microscopy. Selected run products were carbon coated and examined by scanning electron microscopy (SEM) in a Philips XL20 instrument using secondary electron imaging and an accelerating voltage of 20 kV; qualitative elemental analysis was performed by energy dispersive X-ray (EDX) analysis using an Oxford Instruments Si(Li) energy dispersive spectrometer. Three of the samples synthesized in the presence of chromate were also examined as polished thin sections prepared by encapsulating the powder products in epoxy resin. X-ray Powder Diffraction. X-ray powder diffraction (XRPD) data were collected on a Panalytical Expert Pro diffractometer, equipped with an X’celerator position-sensitive detector. All samples were mounted in 1 cm circular cavity holders, and spun at 30 rpm during data collection using Ni filtered Cu radiation. Scans were made between 5 and 75° 2Θ, using a 0.5° divergence slit and step size of 0.017°, counting for 100 s per step. Subsamples were spiked with National Bureau of Standards 640a silicon powder as an internal d-spacing standard. Phases were identified via comparison with patterns in the Powder Diffraction File, set 44, International Centre for Diffraction Data, 2001. Selected samples were ground in an agate mortar and pestle to further reduce particle size and mounted in thin-walled 0.5 mm diameter glass capillaries. The capillaries were spun at 300 rpm and examined in transmission geometry, using Cu KR1 radiation selected by an incident beam hybrid monochromator; the instrument and detector were the same, however, the count time was increased to 500 s per step so as to increase precision. Peak height intensity was measured on background subtracted patterns using the peak search routine of the Bruker Eva Diffrac-Plus software. Unit cell parameters for hydrogarnet were refined with the Bruker AXS program Topas version 2.1, using the LeBail method and space group Ia3d. Transmission Electron Microscopy. Transmission electron microscopy (TEM) was carried out using an FEI CM200 field emission gun instrument operated at 200 kV and fitted with an ultrathin window Oxford Instruments EDX detector and Gatan GIF200 electron energy loss spectroscopy (EELS) imaging filter. The specimens were prepared by dispersing and grinding the samples in methanol in an agate mortar and pestle and pipetting a few drops onto a copper TEM grid coated with a thin, holey amorphous carbon film. Thin sample areas overhanging holes in the support film were imaged and spectroscopically analyzed using EELS and EDX analysis. EEL spectra were recorded in selected area diffraction mode from single-crystal regions of approximately 170 nm in diameter using convergence and collection angles of I4h 3d symmetry at 5 Gpa. Am. Mineral. 2002, 87, 642-647. (24) Sacerdoti, M.; Passaglia, E. The crystal structure of katoite and implications within the hydrogrossular group of minerals. Bull. Mineral. 1985, 108, 1-8. (25) Lager, G. A.; Armbruster, T.; Rotella, F. J.; Rossman, G. R. The OH substitution in garnets: X-ray and neutron-diffraction, infrared and geometric-modeling studies. Am. Mineral. 1989, 74, 840-851. (26) Jappy, T. G.; Glasser, F. P. Synthesis and stability of silica substituted hydrogarnet Ca3Al2Si3-xO12-4x(OH)4x. Adv. Cement Res. 1992, 4, 1-8. (27) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. The United Kingdom Chemical Database Service. J. Chem. Inf. Comput. Sci. 1996, 36, 746-749. (28) Inorganic Crystal Structure Database (ICSD). FIZ Karlsruhe. ICSD-WWW Interface; Hewat, A.; Institut Laue-Langevin, 6, rue Jules Horowitz, BP 156 -38042 Grenoble Cedex 9 - France; http://icsd.ill.fr/icsd. (29) Omotoso, O. E.; Ivey, D. G.; Mikula, R. Hexavalent chromium in tricalcium silicate. Part 1. Quantitative X-ray diffraction analysis of crystalline hydration products. J. Mater. Sci. 1998, 33, 507-513. (30) Palmer, C. D. Precipitates in a Cr(VI)-contaminated concrete. Environ. Sci. Technol. 2000, 34, 4185-4192. (31) Perkins, R. B.; Palmer, C. D. Solubility of Ca6(Al[OH]6)2SO4‚ 26H2O, the chromate analog of ettringite, 5-75 °C. Appl. Geochem. 2000, 15, 1203-1218. (32) Perkins, R. B.; Palmer, C. D. Solubility of chromate hydrocalumite (Ca3Al2O6‚CaCrO4‚12H2O), 5-75.°C. Cem. Concr. Res. 2001, 31, 983-992. (33) Farmer, J. G.; Graham, M. C.; Thomas, R. P.; Licona-Manzur, C.; Paterson, E.; Campbell, C. D.; Geelhoed, J. S.; Lumsdon, D. G.; Meeussen, J. C. L.; Roe, M. J.; Conner, A.; Fallick, A. E.; Bewley, R. J. F. Assessment and modelling of the environmental chemistry and potential for remediative treatment of chromiumcontaminated land. Environ. Geochem. Health 1999, 21, 331337.
Received for review September 14, 2006. Revised manuscript received December 20, 2006. Accepted January 8, 2007. ES0621997
VOL. 41, NO. 6, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1927
